Openclaw Cache Kit

Automatically apply Claude prompt caching optimizations to OpenClaw. Reduces API costs by up to 89% via cacheRetention long, 59-minute heartbeat, cache-ttl p...

What to check before running: - Verify the scripts on a non-production or test host first. They will overwrite $HOME/.openclaw/openclaw.json (a backup is made under $HOME/.openclaw/backups). - Ensure the openclaw CLI and python3 are present (apply.sh will abort otherwise). The skill metadata did not declare these requirements—this is an omission. - Inspect what OpenClaw writes into diagnostics.cacheTrace when includeSystem is true; confirm it does not capture sensitive host or environment data you don’t want logged. Consider changing includeSystem to false if uncertain. - Review the backup file before restoring or rolling out changes at scale. - Run bash scripts/check-savings.sh against a copy of the log file to validate calculations; logs are read locally from ~/.openclaw/logs/cache-trace.jsonl (or the OPENCLAW_CACHE_LOG env override). - Because the manifest omitted declared requirements, proceed cautiously and prefer manual application or a dry-run until you confirm precisely what will be logged and changed.

Type: OpenClaw Skill Name: openclaw-cache-kit Version: 1.1.1 The OpenClaw Cache Kit is designed to optimize OpenClaw's caching mechanisms to reduce API costs. The `scripts/apply.sh` script safely modifies the `openclaw.json` configuration file using an embedded Python script, creates a backup, and restarts the OpenClaw gateway, all actions directly aligned with its stated purpose. The `scripts/check-savings.sh` script reads local diagnostic logs to report cost savings. The `SKILL.md` and `scripts/cron-md-template.md` files provide clear documentation and guidance for optimizing agent behavior, without any evidence of prompt injection for malicious purposes, data exfiltration, unauthorized execution, or persistence mechanisms. All operations are local and transparent.